
# Includes
# os
import os
from time import sleep
import datetime
from datetime import timezone

# websockets and connectivity
from binance.websockets import BinanceSocketManager
from twisted.internet import reactor

# interactive shell
from IPython.core.interactiveshell import InteractiveShell
InteractiveShell.ast_node_interactivity = "all"

# binance
from binance.client import Client
import binance.helpers as btime

# pandas
import pandas as pd
from pandas import DataFrame as df
pd.set_option("display.max_rows", 100, "display.max_columns", None)
pd.set_option('display.float_format', lambda x: '%.5f' % x)


# numpy
import numpy as np

# plotting
from matplotlib import pyplot as plt
%matplotlib inline
import seaborn as sb
from mpl_toolkits.mplot3d import Axes3D
from statsmodels.graphics import tsaplots


# data science utils
from scipy import stats
from scipy import fft
from scipy import signal
import statsmodels.api as sm
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import pingouin as pg

import warnings
warnings.filterwarnings("ignore")


# Load api login data

# open api login file
apifile = open("apilogin.txt", "r");
apilogin = (str(apifile.read()).split("\n"));

# extract key
apikey = apilogin[0].split(":")[1];
# extract secret
apisecret = apilogin[1].split(":")[1];

# check values
"key: " + apikey;
"secret: " + apisecret;

client=Client(api_key=apikey,api_secret=apisecret);


# define trade pair
base_asset = 'ETH'
quote_asset = 'BTC'
trade_pair = base_asset+quote_asset

# define request interval
time_interval = Client.KLINE_INTERVAL_1MINUTE

# define initial start as utc timestamp
start_time = btime.date_to_milliseconds('Dec 1, 2020')

# request initial historical data
candles = client.get_historical_klines(symbol=trade_pair,interval=time_interval, start_str=start_time,limit=1000)


len(candles)
# Verify the last known data point is today
end_unix_timestamp = candles[-1][0]/1000
datetime.datetime.fromtimestamp(end_unix_timestamp)

577020

datetime.datetime(2022, 1, 6, 9, 22)


# Dynamically name headers
kline_headers = ['Open Time UTC', 
    'Open ' + quote_asset, 
    'High ' + quote_asset, 
    'Low ' + quote_asset, 
    'Close ' + quote_asset, 
    'Volume ' + base_asset, 
    'Close Time UTC', 
    'Volume ' + quote_asset, 
    'Number of Trades',
    'Taker Buy Volume ' + base_asset,
    'Taker Buy Volume ' + quote_asset]
# kline_headers


# Create key-value pairs for each column and associated numpy dtype
typedict = {
    kline_headers[0]: np.uint64,
    kline_headers[1]: np.float32,
    kline_headers[2]: np.float32,
    kline_headers[3]: np.float32,
    kline_headers[4]: np.float32,
    kline_headers[5]: np.float32,
    kline_headers[6]: np.uint64,
    kline_headers[7]: np.float32,
    kline_headers[8]: np.float32,
    kline_headers[9]: np.float32,
    kline_headers[10]: np.float32,
}

# Instantiate klines dataframe and assign dtypes via dictionary
klines = df(data=[entry[:-1] for entry in candles], columns=kline_headers).astype(typedict)

# Sort klines dataframe by ascending UTC open
klines.sort_values(by=['Open Time UTC'], ascending=True, inplace=True)

# Assign open UTC as index
klines.index = klines['Open Time UTC']

print('Kline data types:')
klines.dtypes

# klines.head()

Kline data types:

Open Time UTC            uint64
Open BTC                float32
High BTC                float32
Low BTC                 float32
Close BTC               float32
Volume ETH              float32
Close Time UTC           uint64
Volume BTC              float32
Number of Trades        float32
Taker Buy Volume ETH    float32
Taker Buy Volume BTC    float32
dtype: object


# calculate close-open differences for each period
klines['Difference ' + quote_asset] = klines['Close ' + quote_asset]-klines['Open ' + quote_asset]
klines['Percent Change ' + quote_asset] = 100 * klines['Difference ' + quote_asset] / klines['Open ' + quote_asset]
klines.head()


# create Maker-Taker ratio
klines['Maker-Taker Volume ' + quote_asset + ' Ratio'] = ((klines['Volume ' + quote_asset] - klines['Taker Buy Volume ' + quote_asset]) / klines['Volume ' + quote_asset])

# create generic Ratios
klines['Price-Volume ' + quote_asset + ' Ratio'] = klines['Open ' + quote_asset] / klines['Volume ' + quote_asset]
klines['Trade-Volume ' + quote_asset + ' Ratio'] = klines['Number of Trades']/klines['Volume ' + quote_asset]
klines['Price-Taker Buy Volume ' + quote_asset + ' Ratio'] = klines['Open ' + quote_asset] / klines['Taker Buy Volume ' + quote_asset]
klines['Trade-Taker Buy Volume ' + quote_asset + ' Ratio'] = klines['Number of Trades']/klines['Taker Buy Volume ' + quote_asset]


# Differentials lagged by 1 timestep
# so that all diff columns are predictors

klines_diff_1 = klines.diff(axis='index',periods=1).shift(1)
klines_diff_1.columns = ['DIFF 1 ' + colname for colname in klines.columns]
klines_diff_1.drop(columns=[klines_diff_1.columns[0], klines_diff_1.columns[6]], inplace=True)

klines_diff_2 = klines.diff(axis='index',periods=2).shift(1)
klines_diff_2.columns = ['DIFF 2 ' + colname for colname in klines.columns]
klines_diff_2.drop(columns=[klines_diff_2.columns[0], klines_diff_2.columns[6]], inplace=True)

klines_diff_3 = klines.diff(axis='index',periods=3).shift(1)
klines_diff_3.columns = ['DIFF 3 ' + colname for colname in klines.columns]
klines_diff_3.drop(columns=[klines_diff_3.columns[0], klines_diff_3.columns[6]], inplace=True)


klines = klines.join(klines_diff_1,how='inner').join(klines_diff_2,how='inner').join(klines_diff_3,how='inner')
klines.head()


# Verify no NaNs except in first three rows, then remove them in place
klines[klines.isna().any(axis=1)].head()
klines[klines.isna().any(axis=1)].shape
klines.drop(axis='index', labels=klines[klines.isna().any(axis=1)]['Open Time UTC'].values.tolist(),inplace=True)

(164, 66)


# Create df for price data only
price_df = df(klines['Open ' + quote_asset])
price_df.columns=['Price']
price_df.head()

# start with 1st order linear model
time_idx = price_df.index.to_numpy()
price = price_df.Price.to_numpy()
order1_regression = stats.linregress(time_idx,price)
order1_prediction = order1_regression.intercept + order1_regression.slope*time_idx
order1_residual = price-order1_prediction


# Graph true price vs 1st order approximation
plt.figure(figsize=(20,4));
plt.plot(time_idx, price)
plt.plot(time_idx, order1_prediction)
plt.title('Price History (USD) ')
plt.xlabel('UTC Timestamp (s)')
plt.ylabel('Price ($USD)')

# Graph 1st order residual 
plt.figure(figsize=(20,4))
plt.plot(time_idx, order1_residual)
plt.title('')

<Figure size 1440x288 with 0 Axes>

[<matplotlib.lines.Line2D at 0x20af1601dc0>]

[<matplotlib.lines.Line2D at 0x20af161d160>]

Text(0.5, 1.0, 'Price History (USD) ')

Text(0.5, 0, 'UTC Timestamp (s)')

Text(0, 0.5, 'Price ($USD)')

<Figure size 1440x288 with 0 Axes>

[<matplotlib.lines.Line2D at 0x20af29299d0>]

Text(0.5, 1.0, '')


# remove DC component of order 1 residual
order1_residual = order1_residual - np.mean(order1_residual)

# n = signal length
n = price.size

# T = spacing
timestep = 1/(24*60) # unit is days


# Raw signal FFT
price_fft = fft.fft(order1_residual,norm=None)[1:n//2] # fft amplitude of pair price 
fft_freq = 7*fft.fftfreq(n, timestep)[1:n//2] # cycles per week
fft_period = 1/fft_freq # period in weeks

# Plot FFT by frequency
plt.figure(figsize=(24,8));
ax = plt.gca();
ax.plot(fft_freq[:50], (np.abs(price_fft[:50])),linewidth=1);
plt.grid();
plt.title('Discrete Fourier Transform: ' + base_asset + '-' + quote_asset);
plt.ylabel('Price Amplitude');
plt.xlabel('Cycles/week');
plt.gca().xaxis.set_minor_locator(plt.MultipleLocator(1/7));
plt.gca().xaxis.set_major_locator(plt.MultipleLocator(1));

# # Plot FFT by wavelength
# plt.figure(figsize=(24,8))
# ax = plt.gca()
# ax.plot(fft_period, (np.abs(price_fft)),linewidth=1,marker='o')
# plt.grid()
# plt.title('Discrete Fourier Transform: ' + base_asset + '-' + quote_asset)
# plt.ylabel('Price Amplitude');
# plt.xlabel('Period (week)');


fft_data = df(data=[np.abs(price_fft[0:10]),fft_freq[0:10]],index=['Price Amplitude','Cycles per Week']).transpose()
fft_data['Cycle Period (Days)'] = 7/fft_data['Cycles per Week']
fft_data


# Raw signal
order0_residual = price-np.mean(price) # remove DC component
ac_residual = np.correlate(order0_residual, order0_residual, 'same')[n//2+1:] # calucate autocorrelation
ac_residual /= ac_residual[0] # normalize by first value

lag = np.arange(len(ac_residual))*(1/(60*24))/7 # lag in weeks

# plot autocorrelation
plt.figure(figsize=(22,6));
ax = plt.gca();
plt.axhline(y=0, color='k',linewidth=1);
plt.plot(lag, ac_residual, linewidth=3);
plt.grid();
plt.title('Autocorrelation Function: ' + base_asset + '-' + quote_asset);
plt.ylabel('Autocorrelation');
plt.xlabel('Lag (weeks)');
ax.xaxis.set_major_locator(plt.MultipleLocator(1));


# Partial Autocorrelation
plt.figure(figsize=(20,4));
pac = sm.tsa.stattools.pacf(order0_residual, nlags=500, method='ld');
plt.plot(np.arange(len(pac)), pac, linewidth=3);
plt.grid();
plt.title('Partial Autocorrelation Function: ' + base_asset + '-' + quote_asset);
plt.ylabel('Partial Correlation Coefficient');
plt.xlabel('Lag (minutes)');


# Drop time columns as DF in indexed by UTC timestamp
complete_data = klines.drop(['Open Time UTC','Close Time UTC'],axis='columns')

# add lagged data to complete_data
lag1 = complete_data.shift(1).drop(labels=['Open ' + quote_asset, 'High ' + quote_asset, 'Low ' + quote_asset, 'Close ' + quote_asset],axis='columns').drop(labels=[x for x in klines.columns if 'DIFF ' in x], axis='columns').rename(columns={x: 'LAG 1 ' + x for x in klines.columns})
lag2 = complete_data.shift(2).drop(labels=['Open ' + quote_asset, 'High ' + quote_asset, 'Low ' + quote_asset, 'Close ' + quote_asset],axis='columns').drop(labels=[x for x in klines.columns if 'DIFF ' in x], axis='columns').rename(columns={x: 'LAG 2 ' + x for x in klines.columns})
lag3 = complete_data.shift(3).drop(labels=['Open ' + quote_asset, 'High ' + quote_asset, 'Low ' + quote_asset, 'Close ' + quote_asset],axis='columns').drop(labels=[x for x in klines.columns if 'DIFF ' in x], axis='columns').rename(columns={x: 'LAG 3 ' + x for x in klines.columns})

complete_data = complete_data.join(other=[lag1,lag2,lag3],how='inner').dropna()


# df(complete_data.columns)
complete_data.head()
print('Augmented Kline data shape: ')
complete_data.shape

Augmented Kline data shape:

(576853, 100)


# Plot correlation between pairs except for lag
pairplot = sb.pairplot(complete_data[[col for col in complete_data.columns if 'LAG' not in col]].sample(n=1600, axis='index'), diag_kind='kde', kind='scatter',height=2, corner=True,plot_kws=dict(marker='+',linewidth=1))

pairplot.savefig('pairplot_' + trade_pair + '.png',transparent=False)


# Collect pairwise correlation into dataframe
pairwise = pg.pairwise_corr(complete_data.sample(16000), method='spearman', # use spearman correlation to look for monotonically varying pairs
    alternative='two-sided',padjust='bonf').sort_values(by='r',ascending=False)
pairwise


# Clean up pairwise data and copy to new DF 
corr_data = pairwise.sort_values(by='r',ascending=False).drop(columns=['method','alternative','n','power','p-unc'])
corr_data = corr_data[(~corr_data['X'].str.contains(base_asset)) ].set_index('X').drop(columns='p-adjust')


corr_data
plt.figure(figsize=(10,4))
ax = corr_data.r.plot.density(bw_method='scott')
ax.set_title('Smoothed Distribution of\nSpearman Correlation Coefficient for Predictive Pairs');
ax.set_xlabel('r-value');
ax.set_ylabel('PDF');
ax.set_xlim([-1,1]);


# # Groupby each variable and sort/filter to find pairs with |r| > 0.1
# corr_summary = corr_data.groupby(by='Y')
# for x in corr_summary.groups:
#     if ('Percent Change ' + quote_asset in corr_summary.get_group(x).index):
#         corr_summary.get_group(x).sort_values(by='r',ascending=False)[np.abs(corr_summary.get_group(x)['r'])>0.10]


# Find variables that correlate with |r| > 0.1 to predict Maker-Taker Volume Ratio
df([entry[1] for entry in df(complete_data.corr()['Maker-Taker Volume ' + quote_asset + ' Ratio']).transpose()[[colname for colname in complete_data.columns if not 'BTC' in colname]].transpose().sort_values(by='Maker-Taker Volume ' + quote_asset + ' Ratio',ascending=False).iterrows() if np.abs(entry[1][0]) > 0.1] )

# Find variables that correlate with |r| > 0.1 to predict Percent Price Change
df([entry[1] for entry in df(complete_data.corr()['Percent Change ' + quote_asset]).transpose()[[colname for colname in complete_data.columns if not 'BTC' in colname]].transpose().sort_values(by='Percent Change ' + quote_asset,ascending=False).iterrows() if np.abs(entry[1][0] > 0.1) ])


# Create a table of lagged correlations between all variables and Price Change %
lagged_correlation = df.from_dict(
    {x: [complete_data['Percent Change ' + quote_asset].corr(complete_data[x].shift(t)) for t in range(20)] for x in complete_data.columns})
lagged_correlation = lagged_correlation*100

# Filter to find correlations for proportional distributions only,
# since we are only interested in scale-invariant predictors with compact support
x_corr_vol = lagged_correlation[[col for col in lagged_correlation.columns if 'Ratio' in col and 'LAG ' not in col]].transpose().style.background_gradient(cmap=sb.color_palette("coolwarm", as_cmap=True),axis=None).set_precision(1);
x_corr_vol


plt.figure(figsize=(22,5));
plt.plot(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF ' not in col and 'LAG ' not in col)]]);
plt.legend(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF ' not in col and 'LAG ' not in col)]].columns);
plt.axhline(y=0, c='k', lw=1);
plt.gca().xaxis.set_major_locator(plt.MultipleLocator(5))
plt.gca().xaxis.set_minor_locator(plt.MultipleLocator(1))
plt.title('Cross Correlation Between Price Change Percentage and Predictors');
plt.ylabel('r-value (%)');
plt.xlabel('lag (minutes)');
plt.grid('both')


fig, axs = plt.subplots(nrows=3,figsize=(22,6*3))
axs[0].plot(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 1' in col)]],marker='+');
axs[0].legend(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 1' in col)]]);

axs[1].plot(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 2' in col)]]);
axs[1].legend(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 2' in col)]]);

axs[2].plot(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 3' in col)]]);
axs[2].legend(lagged_correlation[[col for col in lagged_correlation.columns if ('Volume ' + quote_asset in col or 'Ratio' in col) and ('DIFF 3' in col)]]);



axs[0].set_title('Cross Correlation Between Price Change Percentage and Predictor Differentials');
for x in axs: x.set_ylabel('r-value (%)');
for x in axs: x.set_xlabel('lag (minutes)');
for x in axs: x.set_xticks(range(0,lagged_correlation.shape[0]));
for x in axs: x.grid('both');
for x in axs: x.axhline(y=0, c='k', lw=1);
for x in axs: x.xaxis.set_major_locator(plt.MultipleLocator(5));
for x in axs: x.xaxis.set_minor_locator(plt.MultipleLocator(1));


InteractiveShell.ast_node_interactivity = "last_expr"
# check normality assumption in Maker-Taker Volume Ratio

# plot KDE, and flip for visual symmetry check
mtvol_sample = complete_data.sample(16000)['Maker-Taker Volume ' + quote_asset + ' Ratio'];

mtvol_desc_stats = mtvol_sample.describe();
mtvol_sample_normalized = (mtvol_sample - mtvol_desc_stats['mean']) / mtvol_desc_stats['std'];

# plot QQ, and flip 180 deg for visual symmetry check
ax = plt.figure(figsize=(12,6));
mtvol_sample_normalized.plot.kde(lw=3,label='KDE Estimate');

(-1.0*mtvol_sample_normalized).plot.kde(lw=3,label='Reversed KDE Estimate');
plt.legend();

plt.title('KDE for the Maker-Taker Volume Ratio');

sm.qqplot(complete_data['Maker-Taker Volume ' + quote_asset + ' Ratio'],
fit=True,markersize=2,label='Maker-Taker QQ').set_size_inches((6,6));

ax = plt.gca();
ax.scatter(-ax.lines[0].get_xdata(),-ax.lines[0].get_ydata(),s=2,c='r',label='$180^\circ$ rotated QQ Plot');
ax.axline([-4,-4],[4,4],c='k',label='Normal Distribution');

plt.title('Maker-Taker Volume Ratio QQ Plot');
plt.legend();

print('Maker-Taker Ratio Kurtosis = ' + str(complete_data['Maker-Taker Volume ' + quote_asset + ' Ratio'].kurtosis()));



# distribution is appproximately normal with slight left-skew and slight negative kurtosis

Maker-Taker Ratio Kurtosis = -0.32462105


# Inspect the 
plot_sample = complete_data.sample(16000);


plt.figure(figsize=(12,8));
# ax = plt.axes(projection='3d')
plt.axhline(0,color='black',lw=1);
plt.axvline(0,color='black',lw=1);

x = (plot_sample['Maker-Taker Volume ' + quote_asset + ' Ratio']).to_numpy()

plt.scatter(x,plot_sample['Percent Change ' + quote_asset],
# c=plot_sample.index, ###
c=plot_sample.index,
s=3,alpha=1,cmap='plasma');
plt.colorbar().set_label('UTC Timestamp (s)');
plt.xlabel('Maker-Taker Volume Ratio');
plt.ylabel('Price Change (%)');
plt.title('Price Change %}' + ' vs ' + 'Maker-Taker Volume Ratio');

#ax=plt.gca();
#plt.xlim([0,1])
#plt.ylim([-2,2])

#ax.yaxis.set_ticks(np.arange(-2,2.2,0.2));

# plt.yscale('symlog',linthresh = 0.001);
# plt.xscale('symlog',linthresh = 0.001);


# # Plot all price difference - x pairs
# plot_sample = complete_data.sample(16000);

# for x in ( x for x in plot_sample.columns if 'Ratio' in x ):
#     plt.figure(figsize=(6,6));
#     # ax = plt.axes(projection='3d')
#     plt.axhline(0,color='black',lw=1);
#     plt.axvline(0,color='black',lw=1);

#     plt.scatter(
#         plot_sample[x],
#         plot_sample['Difference ' + quote_asset],
#         # c=plot_sample.index, ###
#         c=plot_sample.index,cmap='cool',
#         s=3,alpha=1);

#     # plt.colorbar();
#     plt.xlabel(x);
#     plt.ylabel('Price Difference %');
#     plt.title(x + ' vs ' + 'Price Difference %');
    

    #ax=plt.gca();
    #plt.xlim([0,1])
    #plt.ylim([-2,2])

    #ax.yaxis.set_ticks(np.arange(-2,2.2,0.2));

    # plt.yscale('symlog',linthresh = 0.001);
    # plt.xscale('symlog',linthresh = 0.001);
    # plt.yscale('logit');

	Open Time UTC	Open BTC	High BTC	Low BTC	Close BTC	Volume ETH	Close Time UTC	Volume BTC	Number of Trades	Taker Buy Volume ETH	Taker Buy Volume BTC	Difference BTC	Percent Change BTC
Open Time UTC
1606780800000	1606780800000	0.03131	0.03132	0.03125	0.03125	332.94400	1606780859999	10.41869	285.00000	120.16900	3.76137	-0.00006	-0.20121
1606780860000	1606780860000	0.03125	0.03130	0.03125	0.03127	347.59900	1606780919999	10.87070	246.00000	169.12000	5.28886	0.00002	0.06081
1606780920000	1606780920000	0.03126	0.03134	0.03126	0.03133	348.82599	1606780979999	10.92332	232.00000	252.61099	7.91123	0.00007	0.23030
1606780980000	1606780980000	0.03133	0.03137	0.03132	0.03137	377.57199	1606781039999	11.83920	302.00000	284.16000	8.91018	0.00004	0.12127
1606781040000	1606781040000	0.03137	0.03140	0.03134	0.03137	689.13098	1606781099999	21.61878	317.00000	462.52200	14.50975	-0.00000	-0.01274

	Open Time UTC	Open BTC	High BTC	Low BTC	Close BTC	Volume ETH	Close Time UTC	Volume BTC	Number of Trades	Taker Buy Volume ETH	Taker Buy Volume BTC	Difference BTC	Percent Change BTC	Maker-Taker Volume BTC Ratio	Price-Volume BTC Ratio	Trade-Volume BTC Ratio	Price-Taker Buy Volume BTC Ratio	Trade-Taker Buy Volume BTC Ratio	DIFF 1 Open BTC	DIFF 1 High BTC	DIFF 1 Low BTC	DIFF 1 Close BTC	DIFF 1 Volume ETH	DIFF 1 Volume BTC	DIFF 1 Number of Trades	DIFF 1 Taker Buy Volume ETH	DIFF 1 Taker Buy Volume BTC	DIFF 1 Difference BTC	DIFF 1 Percent Change BTC	DIFF 1 Maker-Taker Volume BTC Ratio	DIFF 1 Price-Volume BTC Ratio	DIFF 1 Trade-Volume BTC Ratio	DIFF 1 Price-Taker Buy Volume BTC Ratio	DIFF 1 Trade-Taker Buy Volume BTC Ratio	DIFF 2 Open BTC	DIFF 2 High BTC	DIFF 2 Low BTC	DIFF 2 Close BTC	DIFF 2 Volume ETH	DIFF 2 Volume BTC	DIFF 2 Number of Trades	DIFF 2 Taker Buy Volume ETH	DIFF 2 Taker Buy Volume BTC	DIFF 2 Difference BTC	DIFF 2 Percent Change BTC	DIFF 2 Maker-Taker Volume BTC Ratio	DIFF 2 Price-Volume BTC Ratio	DIFF 2 Trade-Volume BTC Ratio	DIFF 2 Price-Taker Buy Volume BTC Ratio	DIFF 2 Trade-Taker Buy Volume BTC Ratio	DIFF 3 Open BTC	DIFF 3 High BTC	DIFF 3 Low BTC	DIFF 3 Close BTC	DIFF 3 Volume ETH	DIFF 3 Volume BTC	DIFF 3 Number of Trades	DIFF 3 Taker Buy Volume ETH	DIFF 3 Taker Buy Volume BTC	DIFF 3 Difference BTC	DIFF 3 Percent Change BTC	DIFF 3 Maker-Taker Volume BTC Ratio	DIFF 3 Price-Volume BTC Ratio	DIFF 3 Trade-Volume BTC Ratio	DIFF 3 Price-Taker Buy Volume BTC Ratio	DIFF 3 Trade-Taker Buy Volume BTC Ratio
Open Time UTC
1606780800000	1606780800000	0.03131	0.03132	0.03125	0.03125	332.94400	1606780859999	10.41869	285.00000	120.16900	3.76137	-0.00006	-0.20121	0.63898	0.00301	27.35468	0.00832	75.77024	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780860000	1606780860000	0.03125	0.03130	0.03125	0.03127	347.59900	1606780919999	10.87070	246.00000	169.12000	5.28886	0.00002	0.06081	0.51348	0.00287	22.62963	0.00591	46.51286	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780920000	1606780920000	0.03126	0.03134	0.03126	0.03133	348.82599	1606780979999	10.92332	232.00000	252.61099	7.91123	0.00007	0.23030	0.27575	0.00286	21.23897	0.00395	29.32539	-0.00006	-0.00002	0.00000	0.00002	14.65500	0.45201	-39.00000	48.95100	1.52749	0.00008	0.26202	-0.12550	-0.00013	-4.72505	-0.00242	-29.25738	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780980000	1606780980000	0.03133	0.03137	0.03132	0.03137	377.57199	1606781039999	11.83920	302.00000	284.16000	8.91018	0.00004	0.12127	0.24740	0.00265	25.50847	0.00352	33.89380	0.00002	0.00004	0.00001	0.00007	1.22699	0.05261	-14.00000	83.49100	2.62237	0.00005	0.16949	-0.23773	-0.00001	-1.39066	-0.00196	-17.18747	-0.00005	0.00002	0.00001	0.00009	15.88199	0.50463	-53.00000	132.44199	4.14986	0.00013	0.43151	-0.36323	-0.00014	-6.11571	-0.00437	-46.44485	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606781040000	1606781040000	0.03137	0.03140	0.03134	0.03137	689.13098	1606781099999	21.61878	317.00000	462.52200	14.50975	-0.00000	-0.01274	0.32884	0.00145	14.66317	0.00216	21.84738	0.00007	0.00003	0.00006	0.00004	28.74600	0.91588	70.00000	31.54901	0.99895	-0.00003	-0.10903	-0.02835	-0.00022	4.26951	-0.00044	4.56841	0.00009	0.00007	0.00007	0.00011	29.97299	0.96850	56.00000	115.04001	3.62133	0.00002	0.06046	-0.26608	-0.00023	2.87885	-0.00239	-12.61906	0.00002	0.00005	0.00007	0.00013	44.62799	1.42051	17.00000	163.99100	5.14881	0.00010	0.32248	-0.39158	-0.00036	-1.84620	-0.00481	-41.87644

	Open Time UTC	Open BTC	High BTC	Low BTC	Close BTC	Volume ETH	Close Time UTC	Volume BTC	Number of Trades	Taker Buy Volume ETH	Taker Buy Volume BTC	Difference BTC	Percent Change BTC	Maker-Taker Volume BTC Ratio	Price-Volume BTC Ratio	Trade-Volume BTC Ratio	Price-Taker Buy Volume BTC Ratio	Trade-Taker Buy Volume BTC Ratio	DIFF 1 Open BTC	DIFF 1 High BTC	DIFF 1 Low BTC	DIFF 1 Close BTC	DIFF 1 Volume ETH	DIFF 1 Volume BTC	DIFF 1 Number of Trades	DIFF 1 Taker Buy Volume ETH	DIFF 1 Taker Buy Volume BTC	DIFF 1 Difference BTC	DIFF 1 Percent Change BTC	DIFF 1 Maker-Taker Volume BTC Ratio	DIFF 1 Price-Volume BTC Ratio	DIFF 1 Trade-Volume BTC Ratio	DIFF 1 Price-Taker Buy Volume BTC Ratio	DIFF 1 Trade-Taker Buy Volume BTC Ratio	DIFF 2 Open BTC	DIFF 2 High BTC	DIFF 2 Low BTC	DIFF 2 Close BTC	DIFF 2 Volume ETH	DIFF 2 Volume BTC	DIFF 2 Number of Trades	DIFF 2 Taker Buy Volume ETH	DIFF 2 Taker Buy Volume BTC	DIFF 2 Difference BTC	DIFF 2 Percent Change BTC	DIFF 2 Maker-Taker Volume BTC Ratio	DIFF 2 Price-Volume BTC Ratio	DIFF 2 Trade-Volume BTC Ratio	DIFF 2 Price-Taker Buy Volume BTC Ratio	DIFF 2 Trade-Taker Buy Volume BTC Ratio	DIFF 3 Open BTC	DIFF 3 High BTC	DIFF 3 Low BTC	DIFF 3 Close BTC	DIFF 3 Volume ETH	DIFF 3 Volume BTC	DIFF 3 Number of Trades	DIFF 3 Taker Buy Volume ETH	DIFF 3 Taker Buy Volume BTC	DIFF 3 Difference BTC	DIFF 3 Percent Change BTC	DIFF 3 Maker-Taker Volume BTC Ratio	DIFF 3 Price-Volume BTC Ratio	DIFF 3 Trade-Volume BTC Ratio	DIFF 3 Price-Taker Buy Volume BTC Ratio	DIFF 3 Trade-Taker Buy Volume BTC Ratio
Open Time UTC
1606780800000	1606780800000	0.03131	0.03132	0.03125	0.03125	332.94400	1606780859999	10.41869	285.00000	120.16900	3.76137	-0.00006	-0.20121	0.63898	0.00301	27.35468	0.00832	75.77024	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780860000	1606780860000	0.03125	0.03130	0.03125	0.03127	347.59900	1606780919999	10.87070	246.00000	169.12000	5.28886	0.00002	0.06081	0.51348	0.00287	22.62963	0.00591	46.51286	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780920000	1606780920000	0.03126	0.03134	0.03126	0.03133	348.82599	1606780979999	10.92332	232.00000	252.61099	7.91123	0.00007	0.23030	0.27575	0.00286	21.23897	0.00395	29.32539	-0.00006	-0.00002	0.00000	0.00002	14.65500	0.45201	-39.00000	48.95100	1.52749	0.00008	0.26202	-0.12550	-0.00013	-4.72505	-0.00242	-29.25738	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1606780980000	1606780980000	0.03133	0.03137	0.03132	0.03137	377.57199	1606781039999	11.83920	302.00000	284.16000	8.91018	0.00004	0.12127	0.24740	0.00265	25.50847	0.00352	33.89380	0.00002	0.00004	0.00001	0.00007	1.22699	0.05261	-14.00000	83.49100	2.62237	0.00005	0.16949	-0.23773	-0.00001	-1.39066	-0.00196	-17.18747	-0.00005	0.00002	0.00001	0.00009	15.88199	0.50463	-53.00000	132.44199	4.14986	0.00013	0.43151	-0.36323	-0.00014	-6.11571	-0.00437	-46.44485	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1608556800000	1608556800000	0.02686	0.02686	0.02686	0.02686	0.00000	1608556859999	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	NaN	inf	NaN	inf	NaN	-0.00002	-0.00002	-0.00000	0.00002	-191.71600	-5.15132	-129.00000	-77.51800	-2.08311	0.00003	0.13033	-0.24780	0.01576	6.57903	0.01894	-10.53370	-0.00004	-0.00003	-0.00002	0.00000	-166.38701	-4.47386	-109.00000	-15.76800	-0.42478	0.00005	0.16746	-0.47797	0.01528	6.89968	0.00843	-64.09057	-0.00007	-0.00005	-0.00005	-0.00003	-222.13002	-5.97961	-73.00000	-32.16400	-0.86710	0.00005	0.17849	-0.46675	0.01622	17.24460	0.01307	-17.81107

	Price
Open Time UTC
1606781040000	0.03137
1606781100000	0.03137
1606781160000	0.03140
1606781220000	0.03138
1606781280000	0.03132

	Price Amplitude	Cycles per Week	Cycle Period (Days)
0	1382.18137	0.01747	400.59444
1	1550.17458	0.03495	200.29722
2	1205.60322	0.05242	133.53148
3	1778.98307	0.06990	100.14861
4	275.45627	0.08737	80.11889
5	179.75720	0.10484	66.76574
6	546.25039	0.12232	57.22778
7	339.33163	0.13979	50.07431
8	277.48344	0.15727	44.51049
9	347.98698	0.17474	40.05944

	X	Y	method	alternative	n	r	CI95%	p-unc	p-corr	p-adjust	power
100	High BTC	Close BTC	spearman	two-sided	16000	0.99999	[1.0, 1.0]	0.00000	0.00000	bonf	1.00000
197	Low BTC	Close BTC	spearman	two-sided	16000	0.99999	[1.0, 1.0]	0.00000	0.00000	bonf	1.00000
0	Open BTC	High BTC	spearman	two-sided	16000	0.99999	[1.0, 1.0]	0.00000	0.00000	bonf	1.00000
1	Open BTC	Low BTC	spearman	two-sided	16000	0.99999	[1.0, 1.0]	0.00000	0.00000	bonf	1.00000
2	Open BTC	Close BTC	spearman	two-sided	16000	0.99999	[1.0, 1.0]	0.00000	0.00000	bonf	1.00000
...	...	...	...	...	...	...	...	...	...	...	...
4315	DIFF 3 Trade-Taker Buy Volume BTC Ratio	LAG 3 Maker-Taker Volume BTC Ratio	spearman	two-sided	16000	NaN	NaN	NaN	NaN	bonf	NaN
4316	DIFF 3 Trade-Taker Buy Volume BTC Ratio	LAG 3 Price-Volume BTC Ratio	spearman	two-sided	16000	NaN	NaN	NaN	NaN	bonf	NaN
4317	DIFF 3 Trade-Taker Buy Volume BTC Ratio	LAG 3 Trade-Volume BTC Ratio	spearman	two-sided	16000	NaN	NaN	NaN	NaN	bonf	NaN
4318	DIFF 3 Trade-Taker Buy Volume BTC Ratio	LAG 3 Price-Taker Buy Volume BTC Ratio	spearman	two-sided	16000	NaN	NaN	NaN	NaN	bonf	NaN
4319	DIFF 3 Trade-Taker Buy Volume BTC Ratio	LAG 3 Trade-Taker Buy Volume BTC Ratio	spearman	two-sided	16000	NaN	NaN	NaN	NaN	bonf	NaN

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Maker-Taker Volume BTC Ratio	-27.0	-0.4	-0.2	0.0	0.3	-0.1	0.2	0.2	0.1	0.2	-0.2	-0.1	0.1	0.1	0.1	-0.0	-0.0	0.2	0.1	0.1
Price-Volume BTC Ratio	-0.0	-0.0	0.0	0.0	-0.0	-0.1	-0.1	0.0	-0.1	0.0	0.0	-0.0	-0.1	-0.0	-0.0	0.1	0.1	-0.0	0.0	-0.0
Trade-Volume BTC Ratio	0.0	-0.0	0.0	0.0	-0.0	-0.1	-0.1	0.0	-0.1	0.0	0.0	-0.0	-0.1	-0.0	-0.0	0.1	0.0	-0.0	-0.0	-0.0
Price-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
Trade-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 1 Maker-Taker Volume BTC Ratio	-0.2	-0.2	-0.2	0.3	-0.2	0.0	0.1	-0.1	0.3	-0.0	-0.2	0.0	0.0	0.1	0.0	-0.2	0.1	-0.0	0.1	-0.1
DIFF 1 Price-Volume BTC Ratio	-0.0	-0.0	0.0	0.1	-0.0	-0.1	0.1	-0.1	-0.0	0.0	0.1	-0.0	0.0	-0.1	0.0	0.0	-0.0	0.0	0.0	-0.0
DIFF 1 Trade-Volume BTC Ratio	-0.0	-0.0	0.0	0.1	-0.0	-0.1	0.1	-0.1	-0.0	0.0	0.1	-0.0	0.0	-0.1	0.0	0.0	0.0	0.0	0.0	-0.0
DIFF 1 Price-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 1 Trade-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 2 Maker-Taker Volume BTC Ratio	-0.3	-0.3	0.1	0.0	-0.2	0.1	-0.0	0.2	0.2	-0.2	-0.2	0.0	0.1	0.1	-0.1	-0.1	0.1	0.1	-0.0	-0.2
DIFF 2 Price-Volume BTC Ratio	-0.0	0.0	0.1	0.1	-0.1	-0.0	-0.0	-0.1	0.0	0.1	0.0	-0.0	-0.1	-0.1	0.1	0.0	0.0	0.0	-0.0	-0.0
DIFF 2 Trade-Volume BTC Ratio	-0.0	0.0	0.1	0.1	-0.1	-0.0	-0.0	-0.1	0.0	0.1	0.0	-0.0	-0.1	-0.1	0.1	0.0	0.0	0.0	-0.0	-0.1
DIFF 2 Price-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 2 Trade-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 3 Maker-Taker Volume BTC Ratio	-0.5	-0.1	-0.1	0.1	-0.1	0.0	0.3	0.1	0.1	-0.2	-0.2	0.1	0.1	-0.1	-0.1	-0.1	0.2	-0.0	-0.1	-0.2
DIFF 3 Price-Volume BTC Ratio	-0.0	0.1	0.1	-0.0	-0.0	-0.1	-0.0	-0.1	0.1	0.0	0.0	-0.1	-0.1	-0.0	0.0	0.0	0.0	0.0	-0.0	-0.0
DIFF 3 Trade-Volume BTC Ratio	0.0	0.1	0.1	-0.0	-0.0	-0.1	-0.0	-0.0	0.1	0.1	0.0	-0.1	-0.1	-0.0	0.1	0.1	0.0	-0.0	-0.0	-0.0
DIFF 3 Price-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan
DIFF 3 Trade-Taker Buy Volume BTC Ratio	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan